17th conference on "Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes

Size: px
Start display at page:

Download "17th conference on "Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes"

Transcription

1 17th conference on "Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes Budapest (Hungary) 9 May 2016

2 UDINEE project: international platform to evaluate urban dispersion models capabilities to simulate Radiological Dispersion Device Miguel A. Hernández Ceballos 1, Stefano Galmarini 1, Steven Hanna 2, Thomas Mazzola 3, Joseph Chang 4, Roberto Bianconi 5, Roberto Bellasio 5 and UDINEE modelling community 1 European Commission, Joint Research Centre 2 Hanna Consultants, Kennebunkport, ME USA 3 Engility, Lorton, VA, USA 4 Homeland Security Studies and Analysis Institute, Falls Church, VA 5 Enviroware srl, Concorezzo, Italy 2

3 URBAN DISPERSION INTERNATIONAL EVALUATION EXERCISE (UDINEE) Led by the EC-Joint Research Centre (DG-JRC) with the support of the U.S. Defense Threat Reduction Agency (DTRA) 3

4 INDEX 1) Background 2) UDINEE project: Overview 2.1. Objectives and benefits 2.2. Tools: JU2003 and ENSEMBLE 2.3. Participants 2.4. Methodology 3) UDINEE project: Preliminary results 3.1. SF6 0.5 sec time series 3.2. SF6 1 min grid 4) Conclusions 4

5 Background JRC project: Assessment and validation of modelling tools and decision support systems addressing CBRN releases (End date: ) Internationally coordinated model evaluation activities for urban dispersion models 5

6 Objective URBAN DISPERSION INTERNATIONAL EVALUATION EXERCISE (UDINEE) Verify and evaluate the capacities of dispersion models to simulate realistic Radiological Dispersion Device (RDD) in urban environments 6

7 Benefits - assess the real capacity of these systems to respond to emergency situations: timeliness of the prediction, accuracy of the prediction, limitations; - support the use of local models for decision making and policy development; - identify where should be envisage further research to complement missing aspects not yet covered by currently available modelling systems. 7

8 Tools UDINEE is based on: Observational database Joint Urban 2003 (JU2003) field experiment in Oklahoma City Web-based platform ENSEMBLE system PARTICIPANTS 8

9 Participants ESTE CBRN Model name Parallel-Micro-SWIFT-SPRAY (PMSS) microrms / MSS NAME / NAME URBAN EULAG ADREA-HF Canadian Urban Dispersion Modelling System (CUDM) MM5-SCIPUFF/UDM Europe (8 models 6 institutions) Abmerit (Slovak) CEA (France) CNR-ISAC (Italy) Met Office (UK) Institution (Country) National Centre for Nuclear Research (Poland) National Centre for Scientific Research "Demokritos (Greece) North American (2 models 2 institutions) Meteorological Service of Canada (Canada) DTRA (EE UU) 9

10 Tools UDINEE is based on: Observational database Joint Urban 2003 (JU2003) field experiment in Oklahoma City - One of the most comprehensive field campaigns in an urban environment - Urban field experiment with instantaneous non-buoyant releases (by popping a balloon of SF6) at ground levels (1.5 m) with gas/aerosol mixture - similar to RDD - Intensive Operational Periods (IOPs) (Day (5) and night (4)) with large set of monitoring observations 10

11 Overview UDINEE is based on: Web-based platform ENSEMBLE system What is? Platform where several simulations from many models are uploaded together with observations, and a tool for the simultaneous analysis of the results of several models and for the ensemble analysis.

12 Methodology - Definition of the case study (IOPS) - JU2003 data were uploaded 1 - Modellers produce model results - Transform them in ENSEMBLE format - Upload them on the system 2 M1 M2 M3 M4 M5 M.. M9 - Results are ready for consultation 3 - Suites of statistical routines available for quantitative evaluation - Ensemble treatments 4 M1 M2 M.. M9 UDINEE Community 12

13 Deliverables a) Time series a.1) SF6 Sampling sites: 9 (Different locations in each IOP) Reference height: 2 m Time resolutions: 0.5 sec a.2) Meterological parameters Sampling sites: 20 (Same location) Reference height: 2 m Time resolutions: 10 min 13

14 Deliverables b) Grid Grid: horizontal resolution of 5 m (1.4 *1.6 km) and with 57 vertical levels (from 0 to 402 m) Time resolution: 1 min Variables: o Variable 1 - SF6 concentration (pptv) o Variable 2 Average horizontal wind (m / s) o Variable 3 Average vertical wind (m /s) o Variable 4 Average horizontal wind direction (deg) o Variable 5 Turbulent kinetic energy (m2 / s2) 14

15 Preliminary results L05(S1) L17(S8) IOP3 9 th July 2003 L01(S0) L19(S9) L08(S3) L11(S4) L15(S7) L13(S6) Puff Release time Mass (g) Wind speed Wind direction (CST) RELEASE 15

16 Preliminary results SF6 concentration (pptv) Comparison for all values registered in IOP3 1,00E+07 IOP3 - ALL LOCATIONS AND PUFFS P90 1,00E+06 1,00E+05 P10 1,00E+04 1,00E+03 1,00E+02 1,00E+01 1,00E+00 1,00E-01 Obs S0 M1 CA1 FR1 M2 GR1 M3 IT1 M4 PL1 M5 SK1 M6 UK1 M7 UK2 US1 M8 US2 16

17 SF6 concentration (pptv) Preliminary results SF6 concentration (pptv) Comparison for all values registered in the same sampling site during IOP3 1.00E+06 IOP3 - S3 - ALL PUFFS 1.00E+07 IOP3 - S6- ALL PUFFS 1.00E E E E E E E E E E E E E E E-01 S0 CA1 FR1 GR1 IT1 PL1 SK1 UK1 UK2 US1 US2 S0 CA1 FR1 GR1 IT1 PL1 SK1 UK1 UK2 US1 US2 Obs M1 M2 M3 M4 M5 M6 M7 M8 Obs M1 M2 M3 M4 M5 M6 M7 M8 17

18 Preliminary results Comparison of parameters characterizing the time series - Dta= Time difference between Cmax and time when the concentration is greater than or equal to 0.1*Cmax for the first time - Dtd = Time difference between Cmax and time when the concentration is greater than or equal to 0.1*Cmax for the last time 18

19 (sec) Preliminary results (sec) Comparison of parameters characterizing the time series in different sampling sites (sec) (sec) 19

20 Preliminary results b) Grid (SF6 concentration) Horizontal dispersion Horizontal agreement between models; five minutes after the first release 2m Release point Percentage of models in agreement

21 Preliminary results Horizontal dispersion Contribution of each model b) Grid (SF6 concentration) five minutes after the first release 2m Model 1 Release point

22 Preliminary results Horizontal dispersion Contribution of each model b) Grid (SF6 concentration) five minutes after the first release 2m Release point

23 Preliminary results Horizontal dispersion Contribution of each model b) Grid (SF6 concentration) five minutes after the first release 2m Release point

24 Preliminary results Height (m) b) Grid (SF6 concentration) Vertical dispersion average of models five minutes after the first release SF6 concentration (pptv) A B

25 Conclusions - UDINEE is the first multi-model urban dispersion model comparison for RDD releases; - Temporal and spatial (horizontal and vertical) evaluation of models is performed; - The first preliminary results show differences with observations and between model results according to the sampling locations.

26 UDINEE project: international platform to evaluate urban dispersion models capabilities to simulate Radiological Dispersion Device Miguel A. Hernández Ceballos, Stefano Galmarini, Steven Hanna, Thomas Mazzola, Joseph Chang, Roberto Bianconi, Roberto Bellasio and UDINEE modelling community 26